中文摘要：

由于地理位置特殊,景观尺度上的城市森林公园规划实质上是一个多目标的空间优化决策问题。以南京市紫金山森林公园为研究对象,在2005年地类分布格局的基础上,利用CLUE-S模型,预测2011年地类的空间分布,并构建2017年Markov型、结构优化型、城市扩张型3种土地利用情景方案,并对方案进行综合评价。结果表明：1）自然因子,尤其是海拔高度对地类分布起着主要影响;2）基于CLUE-S模型模拟的2011年研究区地类分布格局,与其实际地类分布格局的精度检验指数Kappa值为0.87,模拟效果理想;3）对3种预测情景的综合评价表明,结构优化型是研究区最为理想的发展模式,基于Markov模型所建立的情景规划方案的空间格局很不理想,城市扩张型综合评价性能居中。

英文摘要：

Due to the special geographical location,the planning of urban forest park at landscape scale is essentially a spatial optimization decision making with multiple objective.In this paper,Zijin Mountain Forest Park in Nanjing was taken as the research object.Based on the analysis of land use pattern in 2005and the CLUE-S model,the land use pattern in 2011was predicted and three different land use scenario plannings,namely Markov scenario,structural optimization scenario,urban expansion scenario in 2017were created and comprehensively evaluated.The results showed that：1）natural factors,especially the altitude,played an important role in formation of land use pattern;2）after the comparison between the simulation result of land distribution pattern in Zijin Mountain in 2011and the its practical land distribution pattern,it was found that the value of accuracy evaluation index of Kappa was 0.87,indicating the simulation result was satisfactory;3）it could be seen from the comprehensive evaluation of the three scenarios planning,that the structural optimization scenario was the most ideal scenario,while the Markov scenario was far from ideal and the urban expansion scenario was in the middle.